NO328299B1 - Process for the preparation of 6- (4-chlorophenyl) -2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizin-5-yl-acetic acid and intermediates - Google Patents

Description

The present invention relates to a process for the production of 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizin-5-yl-acetic acid (ML300) and for the production of intermediate products formed by this method.

ML 3000 is a promising cyclooxygenase and 5-lipoxygenase inhibitor and is suitable

- thus for the treatment - of rheumatic diseases and for the preventive treatment of allergic induced diseases, see for example Drugs of the Future 1995, 20 (10): 1007-1009. In this publication, a possible manufacturing route can also be found. Further manufacturing possibilities are described in EP-A-397175, WO95/32970, WO95/32971, WO95/32972, Archiv der Pharmazie 312, 896-907 (1979); and 321, 159-162 (1988), in. Med. Chem. 1994 (37), 1894-1897, Arch. Pharm. With. Chem. 330, 307-312 (1997). In all these syntheses, the structure of the pyrrolizine backbone takes place according to a method shown with the reaction scheme:

The reaction is carried out with methylene chloride, ethanol or diethyl ether. The hydrogen bromide formed by the reaction is captured through the addition of aqueous sodium bicarbonate solution.

The introduction of the acetic acid radical in position 5 can take place through reaction with diazoacetic acid ester, an oxal ester chloride or oxalyl chloride and then saponification or saponification and reduction of the keto group with hydrazine.

In Arch. Pharm. 312, 896-907 (1979), the following turnover is described:

The turnover is carried out in benzene. The COCOCl group is thereby not converted into an acetic acid group, but is reacted with diethylamine.

Impure ML 3000, which is formed by the hydrazine process as a potassium salt and then separated from the reaction mixture by acidification with mineral acid, contains, in addition to potassium salts that are poorly soluble in water, hydrazine, by-products and degradation products (decarboxylation products and dimers) as impurities. This requires additional cleaning operations.

In patent application PCT/EP 01/00852, a method for the production of ML 3000 is specified by reacting 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizine with oxalyl chloride and hydrazine , followed by a special procedure for processing. After the reaction of the pyrrolizine with oxalyl chloride, the product obtained is then treated with hydrazine and an alkali metal hydroxide in an aqueous phase at an elevated temperature, and after finishing the treatment, by adding an ether that is little or not miscible with water, a three-phase system is obtained, and ML 3000 is recovered through acidification of the middle phase. A polymorph ML 3000 is obtained in high yield and with a clean, defined crystalline form.

In total, the synthesis takes place in the steps indicated in the following reaction scheme:

Steps 1 and 2 are known from EP 0 172 371 Al. The reaction of 2,2-dimethyl 1-1,3-propanediol with thionyl chloride is carried out in an inert organic solvent, for example a halogenated hydrocarbon or in an ether, preferably at 0 to 60 °C. The further reaction of the obtained 5,5-dimethyl-1,3,2-dioxathane-2-oxide with sodium cyanide to 4-hydroxy-3,3-dimethylbutyronitrile takes place in DMSO at approx. 80 °C to 120 °C. Stage 1 gives a yield of around 93-99%, and stage 2 gives a yield of around 55-60% with good quality.

For step 3, the reaction with thionyl chloride to 4-chloro-3,3-dimethyl-butyronitrile requires a starting material of high purity. The raw products from steps 1 and 2 must be distilled before further sales.

Also, 4-chloro-3,3-dimethyl-butyronitrile obtained in step 3 must be distilled to achieve the necessary high purity before the subsequent Grignard reaction. With a required purity of 97%, the yield in step 3 is not satisfactory.

Further technical problems arise from the fact that the raw products from reaction steps 1 and 3 are strongly acidic and lead to corrosion of the equipment.

If 4-chloro-3,3-dimelyl-butyronitrile is of the required purity, then the addition of the benzylmagnesium chloride Grignard reagent in step 4 leads to 5-benzyl-3,3-dimethyl-3,4-dihydro-2H-pyrrole and the subsequent cyclization with β-bromo-4-chloroacetophenone in step 5 leads to 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizine in good quality and with a yield of 40-45% over both stages.

The pyrrolizine obtained in step 5 is then, through reaction with oxalyl chloride, followed by reduction with hydrazine in the presence of an alkali metal hydroxide and then acidification, converted to 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3- dihydro-1H-pyrrolizin-5-yl-acetic acid (ML 3000). The yield in this step 6, after cleaning the product, amounts to approx. 62% to 86%.

The known method gives ML 3000 with acceptable purity and acceptable yield, but has some drawbacks, such as problematic chemistry in the second and third steps, an extensive purification of the intermediates is necessary before the further reaction, especially before the Grignard reaction, long service times and corrosion problems in the apparatus when cleaning the strongly acidic extracts from the reaction in steps 1 and 3.

The aim of the present invention is thus to provide a method for the production of 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizin-5-yl-acetic acid (ML 3000 ) for which the disadvantages of the prior art are avoided. With the present method, the aim is in particular to overcome the technical difficulties in steps 1 to 4 of the current synthesis, to avoid the difficult chemistry in the second and third reaction steps of the synthesis, to bypass the Grignard reaction, to increase the total yield, to reduce the residence time and thus overall get a more economical total synthesis.

The present invention thus provides a method for producing the compound of formula I,

characterized by that

a) the compound of formula IV

is converted to the compound of formula III

by that

al) the compound of formula IV is catalytically hydrogenated, or a2) the compound of formula IV is converted into the ketal of formula IVa

where the radicals R, which may be the same or different, are C1-4-alkyl or together form C2-3-alkylene, and the ketal is hydrogenated catalytically, b) the compound of formula III is reacted with o-bromo-4-chloroacetophenone to a compound of formula II

and

c) in connection with formula II an acetic acid radical is introduced.

The object of the present invention is also a method for producing the compound of formula II, characterized in that the compound of formula IV is converted to the compound of formula III by

al) the compound of formula IV is catalytically hydrogenated, or a2) the compound of formula IV is converted to the ketal of formula IVa

where the radicals R, which may be the same or different, are Ci_4-alkyl or together form the C2.3-alkylene,

and the ketal is hydrogenated catalytically, and

b) the compound of formula III is reacted with (o-bromo-4-chloroacetophenone) to a compound of formula II.

The object of the present invention is also a method for producing the compound of formula III, characterized in that the compound of formula IV is converted to the compound of formula III, by

al) the compound of formula IV is hydrogenated catalytically, or

a2) the compound of formula IV is converted into the ketal of formula IVa where the radicals R, which may be the same or different, are C₁-alkyl or together form C₂.3-alkylene,

and the ketal is catalytically hydrogenated.

The introduction of the acetic acid radical in connection with formula II takes place preferably through reaction with oxalyl chloride and reduction of the keto group, preferably with hydrazine and an alkali metal hydroxide,

The preferred method according to the invention can be visualized with the following reaction scheme:

According to the invention, compound IV is preferably produced via the following steps:

1. Preparation of 2-(N-methylaniline)-acrylonitrile (V) from chloroacetaldehyde, N-methylaniline and an alkali metal cyanide, for example potassium cyanide.

v

2. Preparation of 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (IV) through Michael addition of isobutyric nitrile deprotonated with a strong base to the compound of formula V, benzylation of the Michael addition product and hydrolysis of the resulting 2-benzyl-4,4-dimethyl-2-(N-methylaniline)glutaronitrile.

(BnCl = benzyl chloride)

The compounds of formula IV and V, as well as their preparation, are known. Thus, 2-(N-methylaniline)-acrylonitrile (V) is prepared according to H. Ahlbrecht and K. Pfaff, Synthesis. 1980. 413. 2,2-Dimethyl-4-oxo-5-phenyl-valeronitrile (IV) can be prepared according to H. Ahlbrecht and M. Ibe, Synthesis 1985, 421.

However, the method for the production of 2-(N-methylaniline)-acrylonitrile (V) according to the aforementioned method in the literature shows the disadvantage that a complete reaction does not take place either in the reaction with sodium or potassium cyanide or in the basic elimination , that ether is used for extraction and that the product must be purified by distillation, whereby large losses occur through decomposition. According to the invention, it is thus preferred to modify and improve the methods in the literature and combine these in a synthesis in several steps for the preparation of the compound with formula III or formula II or I.

The improvements achieved with the method according to the invention will be explained below. They can be used alone or preferably as a combination. According to the invention, the aforementioned disadvantages are avoided by the starting products being used in second molar ratios and/or N-methylaniline is added instead of chloroacetaldehyde and the reaction components are dosed in, and/or the elimination is carried out in a two-phase system of hydrocarbon/sodium lye with the addition of a phase transfer catalyst, preferably benzyltrielylammonium chloride. Preferred execution conditions must be stated in the following: Chloroacetaldehyde, N-methylaniline and potassium cyanide are used in a molar ratio of from 1.1 to 1.3:1:1.1 to 1.3, in particular approx. 1.2:1:1.2. The addition of N-methylaniline takes place exothermically, and cooling and/or addition speed must be chosen so that the temperature does not exceed 25 °C. For this purpose, N-methylaniline can, for example, be dosed into a mixture of ice and concentrated hydrochloric acid.

Chloroacetaldehyde is preferably added as an aqueous solution to N-methylaniline hydrogen chloride, and through a suitable addition rate and possibly cooling, a temperature of a maximum of 20 °C is maintained. A water solution of calcium cyanide is then added. Through a suitable addition rate and possibly cooling, an upper temperature limit of 20 °C is maintained here as well.

The addition of N-methylaniline, chloroacetaldehyde and potassium cyanide can also take place at a significantly lower temperature, for example below 0 °C. However, temperatures close to the upper limit are preferred, because this allows faster addition and less cooling, without this having any negative effect on yield and product quality.

When the content of N-methylaniline in the formed suspension is below approx. 10%, a solvent which is not miscible with water is added, preferably an aliphatic or aromatic hydrocarbon, especially toluene, and the intermediate product 3-chloro-2-(N-methylaniline)-propionitrile is extracted into the organic phase. The elimination is then carried out in the two-phase system toluene/sodium lye. To accelerate and achieve a complete elimination, a phase transfer catalyst, preferably benzyltriethylammonium chloride, is added. The temperature when adding NaOH should not exceed 15 °C, but after the addition is finished it can rise to room temperature. Using potassium hydroxide instead of sodium hydroxide in the elimination makes the phase separation more difficult.

When the reaction mixture has reached a content of less than 0.5% 3-chloro-2-(N-methylaniline)-propionitrile, the product phase is separated from and possibly washed, for example first with water and then with citric acid water. The organic phase is then dried, for example with magnesium sulphate, and optionally filtered, preferably via a pressure filter. The thus obtained solution of 2-(N-methylaniline)-acrylonitrile in toluene can without problems be stored under nitrogen at -15 °C to -20 °C for further processing. At this temperature, cleavage does not take place. The product is obtained with an excellent yield of approx. 95%, based on methylaniline.

The cyanide-containing waste water and the washing liquids, as well as the rest from the filtration, go to treat the waste water.

The procedure for the preparation of 2,2-dimethyl-5-phenyl-4-oxo*valeronitrile (IV) according to the aforementioned literature reference is followed for benzylation of the Michael addition product at -78 °C. The benzylation is carried out with expensive benzyl bromide, and the hydrolysis and cyanide removal takes place in acetonitrile and requires a reaction time from approx. 40 to 50 hours. The raw product must be distilled. With the present modification of the method for producing the compound IV, distillation of the crude product can be omitted. The purification takes place exclusively through recrystallization. The separation of cyanide takes place in an aqueous/organic system with the addition of a phase transfer catalyst, and by this the reaction time can be significantly shortened. Moreover, the method according to the invention is so far uncomplicated, because it is not necessary to carry out deprotonation and condensation at -78 °C. In addition, no activation with carcinogenic hexamethylphosphoric acid triamide (HMPT) and use of tetrahydrofuran, which is expensive and difficult to dry, is required. Finally, the more affordable benzyl chloride can be used instead of benzyl bromide. Preferred process conditions are stated in the following: Isobutyronitrile is dosed into a solution of a strong base in an inert solvent. As a strong base, for example, sodium amide, sodium urnnaphthalenide and preferably lithium diisopropylamide (LDA) are suitable. The deprotonation preferably takes place in a hydrocarbon, such as ethylbenzene, as a solvent and at temperatures below 10 °C. The compound of formula V is then added, preferably as a toluene solution, and the temperature is also preferably kept below approx. 10 °C. The rate of addition of isobutyronitrile and compound V is chosen accordingly. The preferred reaction temperature for the Michael addition is approx. -10 °C to -20 °C.

When the content of compound V in the reaction mixture has dropped to below approx. 2%, benzyl chloride is added. Preferably, the dosing starts at a lower temperature (approx. -10 °C to approx.

-20 °C), with later heating, for example to approx. 50 °C to 55 °C.

When the content of 2,2-dimethyl-4-(N-methylaniline)-glutaric dinitrile has dropped to below approx. 2%, which takes several hours, the cyanide decomposition is carried out. For this, the benzylated Michael addition product is usually not isolated, but instead converted through acid hydrolysis and liberation of hydrogen cyanide and methylaniline while restoring the carbonyl group in 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (IV). The hydrolysis is carried out after the addition of water, preferably by phase transfer catalysis. As phase transfer catalyst, benzyldiethylammonium chloride or benzyldimethylhexadecylammonium chloride is preferably used. The reaction temperature is usually in the range from approx. 20 °C to approx. 60 °C. By adding a phase transfer catalyst, the reaction time is shortened to approx. 15 to 18 hours. The reaction time can be further shortened if up to approx. 20 vol% methanol and/or use concentrated acid. For example, it is then possible to have reaction times of just 1 hour at approx. 40 °C.

For cyanide decomposition/hydrolysis, a strong mineral acid is added, such as hydrobromic acid or hydrogen chloride acid, and the reaction mixture is preferably allowed to react at an elevated temperature until the content of benzylated Michael addition product has dropped to below approx. 0.5%. The organic phase is then worked up in the usual way and the toluene is distilled off. During the distillation, the temperature should not exceed 50 °C. The distillation residue can then be purified by recrystallization, or prior to recrystallization simultaneous evaporation with isopropanol can be carried out one or more times to remove toluene residues.

The recrystallization can take place in isopropanol, but toluene and mixtures of isopropanol and toluene are also suitable. Preferably, the product is recrystallized from two parts isopropanol/toluene in the ratio 9:1.

Because compound IV has good solubility in isopropanol, cooling is necessary to achieve crystallization, preferably to a temperature of -15°C to -20°C.

Depending on the circumstances, the product obtained may also be contaminated with Michael addition product after the recrystallization. However, such contaminants are not disruptive, as they are easily removed during further sales. Overall, however, the recrystallization has a very good cleaning effect, so that the product is recovered in a very pure form.

As the next reaction step, the obtained 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (IV) is catalytically hydrogenated to 2-benzyl-4,4-dimethyl-1-pyrroline (III). As catalysts, noble metal catalysts such as Pt or Pd are usable. However, Raney catalysts are preferred, particularly Raney-Ni and Raney-Co.

In Arch. Pharm. 299, 518 (1966) describes the preparation of 2-(4-hydroxyphenyl)-4,4-dihydro-3H-pyrrole through hydrogenation of 4-oxo-(4-hydroxyphenyl)butanoic nitrile with Raney nickel. The hydrogenation according to the method according to the invention takes place with an embodiment analogous to the method in the literature, i.e. by using aqueous Raney nickel in alcohols, but only slowly, and at elevated pressure and elevated temperature a clear overhydrogenation is observed.

Attempts were therefore made to shorten the reaction time and to reduce the formation of by-products, particularly as a result of overhydrogenation to pyrrolidine. It turned out that neither raising the hydrogen pressure nor raising the reaction temperature significantly shortened the reaction time, but under certain energetic conditions the proportion of by-products increased, above all partial hydrogenation products, such as oligomeric condensation products and overhydrogenated pyrrolidine.

It has now unexpectedly been found that the quality (purity) of the used starting compound with fonnel IV has a strong influence on both the time course and the by-product distribution during the hydrogenation. The purer the starting material, the simpler and less problematic the course of the reaction. Compound IV is preferably used with a purity of over 90%, in particular over 95% (m/m).

When hydrogenating the nitrile ketone compound with formula IV and using pyrroline with formula III, a tertiary nitrile group is reduced in two partial hydrogenation steps to the neopentylamine group which, during water elimination, spontaneously condenses with the keto group to form the cyclic imino group. The cyclic imino group in the pyrroline can be further hydrogenated to the secondary cyclic amino group in the pyrrolidine. To prevent this, Raney nickel is used as a catalyst, but not as usual in aqueous form, on the other hand as essentially anhydrous. As a solvent, it has been shown that toluene. and in particular mixtures of toluene and Ci.4 alcohols such as methanol, ethanol, isopropanol, for example toluene/methanol in a volume ratio of 8:2 to 6:4, are most suitable.

A further possibility to suppress overhydrogenation consists in introducing an acetal (ketal) protecting group for the keto group of the nitrile ketone, thereby obtaining the compound of formula IVa

where the radical R, which may be the same or different, is a C1-4 alkyl radical or together forms a C2-3 alkylene radical. The hydrogenation of the tertiary nitrile group in the neopentylamine group can then be carried out under less selective conditions, whereby a compound of formula IVb is obtained

where the radical R has the meaning indicated above. Under conditions for ketal cleavage in acidic media, for example in dilute mineral acids. cyclization to pyrroline also occurs at the same time. After the acidic aqueous solution of the pyrroline salt has been made basic, the free pyrroline base is obtained which is separated with an organic solvent which is not miscible with water, and after the removal of this solvent can be obtained in its pure form.

Preferred reaction conditions for the direct preparation of compound III through hydrogenation of the compound with fonnel IV. must be stated in the following: When a mixture of toluene and methanol is used as solvent, preferably approx. 8 to 12 parts by volume toluene/methanol per weight part of compound V. then the reaction temperature is usually approx. 50-60 °C. If the hydrogenation takes place in pure toluene. the temperature is kept somewhat lower, for example at 20-30 °C, to prevent overhydrogenation. The hydrogen pressure is usually approx. 4-6 bars.

Before the reaction, the Raney nickel used is dried, for example by slurrying one or more times in absolute methanol, or by azeotropic distillation.

If the reaction stops before the theoretical amount of hydrogen has been taken up, the reaction mixture can be azeotropically distilled and fresh solvent added. Fresh Raney nickel may also be added and subjected to azeotropic distillation to remove water. The reaction duration is usually approx. 3 to 4 hours.

The Raney nickel is then sedimented and the reaction solution above is filtered. The catalyst can optionally be used again for further hydrogenation. The solvent is distilled off from the reaction solution. The product can be purified through salt formation, for example through the formation of hydrogen chloride, and the compound with fonnel IV is released with the aid of a base, for example ammonia, and re-extraction.

Alternatively, only part of the solvent can be distilled off, for example methanol when a solvent mixture of toluene/methanol is used. In this case, the distillation residue is preferably first washed with water, and after separating from the water phase, the product can be purified as described above.

During the course of the reaction, the hydrogenation, in particular by using anhydrous Raney nickel as catalyst and toluene or a mixture of toluene and methanol as solvent, can be greatly accelerated and side reactions can be kept within certain limits.

The preferred reaction conditions for recovering the compound of formula III through hydrogenation of cyclic or acyclic acetal (ketal) intermediates shall be stated as follows: The nitrile ketone of formula IV is, in a solvent forming an azeotrope with water, converted to the ketal with an alcohol in the presence of an acid catalyst, or the conversion of the ketone to ketal is carried out in an alcohol in the presence of equivalent amounts of an acetal or ketal of a low-boiling aldehyde or ketone. Suitable alcohols for ketal formation are C1.4-alkanols, such as methanol, ethanol or glycol, 1,3-propylene glycol, etc. Solvents that form an azeotrope with water are, for example, toluene, xylene, cyclohexane, etc.

A preferred embodiment is, for example, conversion to an oxolane derivative in toluene by using ethylene glycol in the presence of an acid, such as toluenesulfonic acid, under reflux conditions and with removal of the water from the reaction mixture, for example by using a water separator. A further preferred embodiment is to convert the dimethyl ketal into methanol using 1,1-dimethoxyethane in the presence of pyridine tosylate at approx.

40 to 60 °C. Next, the ketal is treated by being washed with alkali and hydrogenated in the presence of a hydrogenation catalyst. A particularly preferred embodiment is hydrogenation of the dioxolane derivatives in the presence of anhydrous Raney nickel at 5 to 50 bar hydrogen pressure and at room temperature to 70°C in an alcoholic solvent, such as methanol, or in an aromatic solvent, such as toluene. The compound with formula III is obtained after the catalyst has been filtered off by stirring the amino ketals obtained as hydrogenation product from the organic phase into an aqueous, dilute mineral acid. Both the cleavage of the ketals and the formation of the cyclic imine occur completely at room temperature, usually within 30 minutes to 1 lime. By making the aqueous product solution basic to pH 9 to IT 1, the cyclic imine III can be recovered in very pure form.

Then 2-benzyl-4,4-dimethyl-1-pyrroline of formula IV is cyclized with α>-bromo-4-chloroacetophenone to 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3 -dihydro-1H-pyrrolizine with formula III. The reaction is known from the initially mentioned known technique, co-bromo-4-chloroacetophenone can be obtained as described for example in Bull. Soc. Chim. Fr. 21, 69 (1899).

The reaction of the compound having formula III with co-bromo-4-chloroacetophenone usually takes place in a polar organic solvent. Suitable polar organic solvents are particularly Ci_4 alcohols such as methanol, ethanol or isopropanol, or ethers such as diethyl ether, tetrahydrofuran (THF), or dioxane. According to the present invention, methanol is particularly preferred as a solvent. The reaction components can be used in equimolar amounts. However, it is preferred to use co-bromo-4-chloroacetophenone in excess, for example in an excess of from 10 mol% to 40 mol%.

To capture hydrogen bromide that is released during the reaction, one works in the presence of a base. An inorganic base is preferably used, in particular an alkali metal hydrogen carbonate or alkali metal carbonate, where sodium and potassium compounds are particularly preferred. The inorganic base can be used in the form of an aqueous solution. However, it has proven to be particularly preferred to use the inorganic base in solid form. This makes it easier to separate from the inorganic reaction products and reduces the range of by-products. The inorganic base can be used in equimolar amounts based on the amount of hydrogen bromide released. However, the inorganic base is suitably used in excess, for example in an excess of up to 1.8 equivalents, preferably approx. 1.4 equivalents. Moreover, it is appropriate to carry out the turnover in the absence of light. The reaction temperature can be varied within a wide range and is preferably in a range from 0 °C to 50 °C, particularly preferably from approx. 18 °C to 25. The turnover is complete after approx. 17 to 20 hours.

The crude product obtained with fonnel II is separated, for example by centrifugation, and purified in the usual way by removing the accompanying mineral substances. To achieve this, the raw product is preferably filled in distilled water, for example at 40 to 45°C, and treated for 1 to 2 hours. In this way, the compound with fonnel II is obtained in an average yield of 58% and with a purity of at least 97%. The content of the isomer with the 4-chlorophenyl group in position 5 is below 2%, the content of co-bromo-4-chloroacetophenone is below 0.1% and the content of mineral byproducts below 0.5%.

For the preparation of ML 3000 (I), an acetic acid side chain is introduced into the 5-position of the compound with fonnel II. This is preferably done by reacting the compound with formula II with oxalyl chloride and then carrying out reduction with hydrazine and an alkali metal hydroxide. The reaction is described, for example, in WO95/32971, example 5C, and in PCT/EP01/00852. Different ways have been described to purify the reaction product. According to WO95/32971, water is added to the reaction mixture, acidified and the separated carboxylic acid is taken up in diethyl alcohol. The product is purified by stirring the ethereal solution and allowing it to stand for some time over a desiccant such as anhydrous sodium sulphate or magnesium sulphate, then water saturated with sulphate is filtered - from and finally the ether - is evaporated off through heating. The substance which crystallizes from the mother liquor during concentration is collected and dried. In this method of isolation and purification, some degradation products are also formed in the purification step and during drying, so that a further extensive purification of ML 3000 is necessary, for example by recrystallization, in order to achieve pharmaceutical quality.

In the alternative method for purification, after the reduction with hydrazine and an alkali metal hydroxide, an ether and water are added to the reaction mixture, optionally at a higher temperature. Preferably, an ether is used which has limited miscibility with water, for example diethyl ether or methyl-t-butyl ether. By adding the ether, a three-phase system is formed where the middle phase is the product phase, and this essentially consists of the salt of ML 3000 and the alkali metal hydroxide used in the reaction. The upper phase is the ether phase where the organic contaminants are located, and the lower phase is a strongly alkaline aqueous phase containing the inorganic constituents.

The phases are separated and the middle phase is added to a mixture of water and the ether which has limited miscibility with water, then acidification is carried out with an inorganic or organic acid. ML 3000 is then dissolved in the ether phase.

The recovery of ML 3000 from the ether phase can, for example, take place by evaporating the ether and crystallizing ML 3000 from ethyl acetate or isopropanol. A solvate is thereby obtained with the ratio of one molecule of diethyl ether to two molecules of ML 3000, or with one molecule of ethyl acetate to two molecules of ML 3000.

A substantially solvent-free crystal modification of ML 3000 is achieved when a hydrocarbon with a higher boiling point than the ether is added to the ether phase, the ether is at least partially distilled off, and the precipitated ML 3000 in solid, crystalline form is separated from the mother liquor in the usual way. A straight-chain or branched aliphatic C6-12 hydrocarbon, for example n-hexane, n-heptane, cyclohexane, cycloheptane, etc., is particularly useful as a hydrocarbon.

The following examples explain the invention in more detail.

Example 1

A) 2- (N-methylaniline) acrylonitrile (approx. 50% solution in toluene)

Into a 250 L enamel reactor was charged concentrated HCl (32%, 22.33 kg) and ice (32.6 kg). N-methylaniline (17.39 kg, 162.2 mol) was added under water cooling so that the temperature was not allowed to rise above 25 °C (30 min). The greenish-yellow solution was stirred for 5-10 min at 15-20 °C (30 min). After the mixing time, the reaction mixture was stirred for a further 5-10 min at 15-20 °C, and then at this temperature a solution of potassium cyanide (12.7 kg, 195.1 mol) in water (19.5 kg) was added. By means of water cooling, the addition was controlled so that the temperature did not rise above 20 °C (1 h). At a temperature of 18-23 °C, it was stirred for 110-130 min. The gas chromatographic sample showed a content of less than 10% methylaniline. The reaction mixture was then added to toluene (25.7 kg) and finally, with stirring, concentrated hydrochloric acid (32%, 9.3 kg) and stirred for 5-10 min at room temperature.

The hydrocyanic acid that came out of the apparatus was kept - back in an adsorber filled with concentrated NaOH.

With the stirrer turned off, the water phase (114 kg, cyanide waste water 1) was allowed to settle and was then transferred to a closed system in a tank for further treatment.

To the blue colored organic phase was added benzyltriethylammonium chloride (0.3 kg) and cooled to -5°C to 0°C. When this internal temperature was reached, sodium hydroxide liquor (30%, 32.6 kg) was allowed to flow in so that the internal temperature did not exceed 15 °C (30 min). After complete addition, the reaction mixture was allowed to warm to room temperature and was stirred for an additional 50-70 min.

GC analysis of a sample showed a content of less than 0.5% of the intermediate 3-chloro-2-(N-methylaniline)-propionitrile. When this value was reached, it was washed with water (40.7 kg): the water was added, the two-phase mixture was stirred for 5-10 min, then the water phase (79 kg, cyanide waste water 2) was allowed to rest and was transferred to the tank ( to cyanide waste water 1).

The organic phase was again washed in the same way with water (40.7 kg) which had been acidified with citric acid (0.81 kg).

This citric acid water phase (45 kg, cyanide waste water 3) was combined with the other cyanide waste water. The organic phase was dried at room temperature over magnesium sulfate (3.8 kg) for 10-20 min. A Karl-Fischer titration gave a water content of less than 0.2%. This toluene solution (50-52 kg) was filtered through a pressure filter and withdrawn for use in the next step. The filter residue (4.8 kg) was combined with the cyanide waste water. This cyanide waste water was taken to a treatment of the waste water. The solution of 2-(N-methylaniline)-acrylonitrile (53.86 kg) which is unstable at room temperature, was stored under nitrogen at -15°C to -20°C for further workup. A 50 ml sample was taken to determine the content. With 30 ml of this sample, the dry residue was determined by evaporating the toluene in vacuum at a maximum of 70 °C. To determine the content, the integral areas under a 1 H-NMR spectrum of the sample in chloroform and a GC analysis were used. According to 1 H-NMR, the content of 2-(N-methylaniline)-acrylonitrile in the solution was 45.54%. The yield was thus 95.4% based on the methylaniline used.

The treatment of the cyanide waste water takes place with concentrated H202 and 30% NaOH at pH 10-12 until a residual content of cyanide is below 30 mg/kg (<30 ppm).

B) 2. 2-dimethyl-4-oxo-5-phenylvaleronitrile

A solution of lithium diisopropylamide in THF/n-hexane (LDA solution 25.1% by weight, approx. 2 M, 80.7 kg, 188.7 mol) was filled in dry apparatus flushed with protective gas (steel vessel 250 1) under nitrogen and cooled using salt water cooling to -15 °C to -20 °C. During cooling, isobutyronitrile (11.4 kg, 165 mol) was metered in so that the internal temperature did not exceed -10 °C. After the addition was finished, the feed vessel was flushed (45 min) with toluene (2 kg).

The reaction mixture was stirred at temperatures between -10 °C and -20 °C for 55-65 min. Then the toluene solution of 2-(N-methylaniline)-acrylonitrile (47.1%, 52.8 kg, 157.2 mol) was dosed in so that the internal temperature did not exceed -10 °C while cooling with brine at -20 °C C (90 min). The feed vessel and feed line were flushed with toluene (5.0 kg). The red-brown reaction mixture was stirred at -10°C to -20°C for 60-90 min. A gas chromatographic analysis showed that the content of the starting material (2-(N-methylaniline)-acrylonitrile) was below 2%.

With the cooling off, at the start of -10 °C to -20 °C benzyl chloride (23.9 kg, 188.8 mol) was added, and the internal temperature was allowed to rise to 5 °C. If this temperature was exceeded, the reaction was carried out with water cooling. When an internal temperature of 15 °C is reached, the mixture was heated at a heating rate of 20 °C/h to 50 °C, while benzyl chloride was metered in. The time required for the addition was 2.5 h.

The reaction mixture was kept at 50-55 °C for 3-4 h, and the content of 2,2-dimethyl-4-(N-methylaniline)glutaric acid dinitrile in a gas chromatographic sample was then below 2%.

The batch was then cooled below 25°C and transferred to a vessel containing a three-phase mixture of ice (22.6 kg), water (45.2 kg) and toluene (22.6 kg) (10 min). Toluene (14 kg) was used for backwashing. This toluene/water phase mixture was then heated to 35-40 °C and the phases allowed to separate. The clear overlying phase was removed (water phase, 75 kg) and the intermediate layer with the organic product phase was left.

Benzytriethylammonium chloride (3.4 kg) and ice (34.7 kg) were first added to the organic phase, and then hydrobromic acid (48%, 69.4 kg, 411.6 mol) was added at 0-15 °C within 10 min. The temperature in the batch thereby rose to approx. 50 °C and it was powered by hydrocyanic acid which was retained in an absorber filled with caustic soda (32%). After stirring for 6 h at 50-60 °C, a sample of the red-brown reaction mixture was taken. According to GC analysis (GC = gas chromatography), the content of 4-benzyl-2,2-dimethyl-4-(N-methylaniline)-glutaric acid dinitrile in the mixture must be below 0.5%.

If this condition was met, the phases were allowed to rest at an internal temperature of below 60 °C for 10-15 min, and the hydrogen cyanide-containing HBr-acidic dark water phase (cyanide waste water 1, 90-110 kg) was transferred to a tight , closed tank. The similarly dark-colored organic phase was cooled to below 30 °C and then stirred out with a mixture of water (22.5 kg) and caustic soda (30%, 2.5 kg) during 5-10 min at 15-25 °C. The alkaline water phase (pH 10-14) with a distinctly lighter color was drawn off and into a tank for post-treatment (cyanide waste water 2.25 kg). Then the organic phase was stirred with water (25 kg) at 15-25 °C for 10-15 min, and after 10-15 min the water phase was completely separated as alkaline cyanide waste water 2 (25 kg). The pH value in this washing liquid should be 7-9.

The toluene phase was transferred to a distillation apparatus, vessels and inlet connections were flushed with toluene (5 kg). Toluenel was completely distilled off in vacuum to a maximum of 50 °C (distillate 1a, 110-120 kg). The distillation residue was taken up in isopropanol (22.7 kg) and the solvents were completely distilled off (distillate lb, 23 kg) to a maximum internal temperature of 60 °C. The azeotropic distillation with isopropanol (22.7 kg) was repeated again in the same way (distillate 1c, 23 kg).

The residue from the azeotropic distillation was taken up in isopropanol (16 kg) at 25-30 °C and added to a mixture of isopropanol (8.0 kg) and heptane (16 kg), and to control the crystallization, crystal seeds of 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (0.05 kg). Supply vessels and connecting lines were flushed with isopropanol (2.0 kg). The crystal suspension was cooled to -15 °C to -20 °C and stirred for at least 2 h but not more than 16 h. The crystal mass was suctioned off and resuspended in a pre-cooled mixture of isopropanol (8 kg) and heptane (8 kg) at -15 °C to -20 °C for several minutes, and again suctioned off. 26.7 kg of 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile were obtained as moist raw material, in addition to a total of 66.9 kg of mother liquor. The crystals were dried in vacuum at 30-35 °C, and after drying there was 22.3 kg (70.6%) of a product with a purity of over 90% according to the GC analysis.

C) Purification of 2, 2- dimethyl- 4- oxo- 5- phenylvaleronitrile

In a 250 L enamel reactor, 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (85-90%), 22.3 kg, 110.8 mol) was suspended in a mixture of isopropanol (40.0 kg) and toluene (4.4 kg) and on heating this mixture to 50-55 °C with stirring it was completely dissolved. The solution was then cooled to 25-30 °C and poured into a stirred filter drier filled with isopropanol (5 kg), and to the solution was added crystal seeds of crystalline 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (0 .05 kg), and then cooled slowly to 5-10 °C. It was stirred until a thick crystal suspension had formed. It was then cooled to -15 °C to -20 °C and stirred for at least 2 h or overnight at this temperature.

The product was filtered off and washed twice with isopropanol (each time 4.8 kg) which had previously been cooled to -15°C to -20°C. The moist crystal mass (26.6 kg) was dried in vacuum at 30-35 °C, and 16.6 kg of product (74.4% yield) with a purity of 96.1% (GC analysis) was obtained. The mother liquor (53.2 kg) was discarded.

D) 2-benzyl-4,4-dimethyl-1-pyrroline

Raney nickel (7.7 kg) which had been freed from the aqueous overlying phase by decantation was placed in a 250 L steel autoclave covered with nitrogen gas and then suspended for 15 min in methanol (67 kg). After the stirring was stopped, the Raney nickel was allowed to rest for 15-30 min and the overlying methanol was mixed with nitrogen and by means of a dip tube forced through a pressure filter coated with "Dicalite". The catalyst was covered with a solution of 2,2-dimethyl-4-oxo-5-phenyl-valeronitrile (13.2 kg) in toluene (92.4 kg) at 15-20 °C and mixed with methanol which was used for back-flushing of the supply container for loluen solution. The apparatus was filled three times with nitrogen to 3 bar and the pressure relieved, thereby displacing atmospheric oxygen. It was then flushed three times with hydrogen at 1 bar and finally the hydrogen pressure was increased to 4.5-5.5 bar. The hydrogenation was started at 5.0 bar and 55-60 °C with stirring. The hydrogen absorption stopped after approx. 3 h, and at this time 3.3 m3 of hydrogen had been taken up. The reaction mixture was cooled to 15-20 °C, the stirring stopped and the hydrogen overpressure relieved. The apparatus was flushed four times with nitrogen, and a sample was taken to check the reaction. The sum of unconverted starting material and overhydrogenated by-product must not exceed 10%. If the sample shows the required result, the reaction solution is filtered through a pressure filter coated with Dicalite (0.5 kg) until it is clear. Apparatus and filter residue were rinsed with methanol (10 kg) and then methanol was distilled off from the reaction solution at an internal temperature of 75-80 °C. The distillation residue was cooled to 20-30 °C and washed with water (49.5 kg). The two-phase mixture was stirred for 5-10 min, after which it was allowed to stand for 20-30 min for phase separation and then the water phase (47-51 kg) was removed. At 15-20 °C, ice (44 kg) and water (44 kg) and finally concentrated hydrochloric acid (32%, 17.7 kg) were added to the organic phase and stirred for 5-10 min. The HCl-acidic water phase had a pH value of 1-2. Both phases were allowed to rest (10-20 min) and the aqueous pyrroline extraction phase was separated. To this HCl-acidic aqueous product phase was added "Mannite", water which had been used to backflush the line (5.7 kg) and toluene (86.9 kg). During cooling, ammonia solution (24%, 17.7 kg) was added at a maximum of 25 °C. The pH value in the water phase of the phase mixture should be 9-11. The two-phase mixture was stirred for 5-10 min. The phases were then allowed to rest and the water phase was separated. The toluene phase was transferred through backwashing with toluene (5.5 kg) to a distillation apparatus, and the toluene was completely distilled off under vacuum at an internal temperature that did not exceed 50 °C. The recovered toluene distillate can be reused for extraction. The content of pyrroline was determined in an aliquot of the toluene phase (50 g), where the dry residue was determined first through complete evaporation of the toluene in vacuum. This dryer had a content of 70% of the desired 2-benzyl-4,4-dimethyl-1-pyrroline according to GC.

From 100.7 kg of product solution with a dry weight proportion of 13.74% in the 50 g sample and a GC content of 74.1%, a yield of 54.7 mol of 2-benzyl-4,4-dimethyl- l-pyrroline. Based on the oxovaleronitrile used, this gave a yield of 84%.

E) 6-(4-chlorophenyl)-2.2-dimethyl-4-phenyl-2,3-dihydro-1H-pyrrolizine

For the direct subsequent synthesis with ring closure of pyrrolizine, α-bromo-4-chloroacetophenone was used in 10 mol% excess (60.2 mol) and sodium bicarbonate in 36 mol% excess (74.4 mol) in relation to the determined pyrroline ( 54.7 mol).

The distillation residue from step D was mixed at 15-20 °C with methanol (49 kg) then with sodium bicarbonate (6.25 kg) and finally, under cooling, co-bromo-4-chloroacetophenone (14.06 kg). The resulting pale yellow thin suspension was stirred in the absence of light for 17-20 h at 18-25 °C. The suspension was centrifuged and the centrifugate was washed with methanol (11 kg) in two portions.

The methanol mother liquor and melanol wash solution were discarded. 16.5-18.5 kg of moist crude product was obtained, which was slurried in water (88 kg) and stirred at 40-45 °C for 1-2 h. The crude product cleaned of mineral impurities was centrifuged and washed with water (22 kg ) in two portions. The yield of moist raw product was 14-16 kg. The aqueous mother liquor and the aqueous wash phase were discarded.

The crude product was dried in vacuum at 35-40 °C. During drying, the weight amount was reduced to 12.5-13.5 kg (38.4 mol - 41.95 mol) 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H -pyrrolizine with a content of 97.3% (HPLC). This corresponds to a yield of 71.0-76.7% based on pyrroline formed during the hydrogenation, and a yield of 59-64% based on the oxo-valeronitrile used during the hydrogenation. The content of the isomer 5-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizine was less than 2%, the content of co-bromo-4-chloroacetophenone less than 0.1 % and the content of mineral impurities of less than 0.5%

(ash determination).

Example 2

Steps A) to D) in Example 1 were carried out.

E) 2-Benzyl-2-(2-cyano-2-methyl-propyl) 1,3-dioxolane

Oxovaleronitrile (50 g, 0.25 mol) was mixed with ethylene glycol (75 g, 1.21 mol) and p-toluenesulfonic acid (9.2 g, 0.048 mol) in toluene (300 mL, 260.1 g, 2.82 mol), and the reaction mixture was slowly heated to boiling (2.5 h). After a subsequent 2 h reflux, the batch was checked with GC. During the heating and refluxing phase, the toluene was distilled off and replaced with dry solvent (185.3 g). For further processing, the batch was allowed to cool under dry N2. For work-up, the toluene solution of the crude product was extracted with ice-cold caustic soda (25 g, 0.625 mol NaOH in 150 g ice) and the phases were separated. The organic phase was dried with anhydrous magnesium sulfate (MW 120.37, 50 g, 0.4 mol). After the filtration, 245 g of filtrate was obtained.

F) 2-benzyl-4,4-dimethyl-1-pyrroline

The crude solution of dioxolane obtained under E) was charged into a 1 L autoclave and then 20 g of Raney nickel B113W (MW 58.71, 0.34 mol) which had previously been extracted three times with anhydrous methanol, together with 71.1 g of toluene. Through pressurization with nitrogen and subsequent relaxation three times, air oxygen was displaced from the autoclave. The hydrogenation started after hydrogen had been supplied and vented three times in succession, and a hydrogenation pressure of 48 bar and a mantle temperature of the autoclave of 63 °C (duration 3 h) had been set. The hydrogenation in the autoclave of 1 1 required top-up of hydrogen to the initial pressure value after 3 h (internal pressure 23 bar) and after a further 18 g

(internal pressure 17 bar). After a hydrogenation duration of a total of 26.5 h, the whole was allowed to cool and the reaction product was filtered over Decalit.

The cleavage of the acetals occurred directly in connection with this by taking up the crude product in dilute hydrochloric acid (HC1 32%, 50 g, 0.43 mol in H20, 200 g) and stirring for 1 h at 30 °C. The overlying organic phase (toluene phase) was drawn off and at 0 °C to -5 °C the aqueous phase was made basic with aqueous concentrated ammonia (25%, 50 g, 0.73 mol) to a pH of 9 to 10. The precipitated pyrroline was taken up in diethyl ether (200 g) and separated. After the ether had evaporated in vacuo, 32.1 g of product remained. 2-benzyl-4,4-dimethyl-1-pyrroline was obtained in a yield of 69% and with a purity of 92.6% (GC).

If the dioxolane was purified by distillation (92%, GC) before it was used in the hydrogenation, a higher hydrogenation rate was achieved during the hydrogenation at lower pressure (5 bar) and lower temperature. The obtained pyrroline had a purity of 94-98% (GC).

Example 3

Production of ML 3000:

A) 5-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizine

In a reactor (500 1) was filled 17.9 kg (95.5 mol) of 2-benzyl-4,4-dimethyl-1-pyrroline (based on pyrroline compound content) prepared according to example 1 or 2, 29, 7 kg (127.2 mol, 1.33 equivalents) of o-bromo-4-chloroacetophenone and 226.6 kg of methanol. After addition of 12.7 kg (151.2 mol, 1.58 equivalents) of sodium bicarbonate in the dark at 17-24°C with stirring, a beige suspension was formed. The reaction was allowed to proceed until the residual content of pyrroline compound in the mixture was < 5%. After 17 h, a sample was taken and the content of pyrroline compound determined by gas chromatography. The analysis showed a content of 2%. The suspension was then centrifuged at a temperature of 18-22°C, and the solid obtained by the centrifugation was washed with 14.4 kg of methanol in two portions. The still moist, slightly yellow product weighed 25.8 kg.

The still moist crude product (25.8 kg) was suspended in 150 kg of water, then within 15 minutes the water was brought up to a temperature of 50-60 °C and stirred for 40 min at this temperature. The suspension was cooled to 40°C (40 min) and centrifuged, and the pale yellow crystalline solid obtained by the centrifugation was washed with 27 kg of water in two portions. The product was dried in vacuum at 50-60 °C for 12-24 h. 18.6 kg of 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H- pyrrolizine with an ash content of 0.33% and an isomer content of 5-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizine of 1.0%.

B) 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3-dihydro-1H-pyrrolizin-5-yl-acetic acid (ML 3000)

Into a 250 L reactor that had been evacuated and filled with N2 three times, 11.5 kg (35.7 mol) of 6-(4-chlorophenyl)-2,2-dimethyl-7-phenyl-2,3- dihydro-1H-pyrrolizine in 60 kg of tetrahydrofuran (THF). Under a 0.5 bar nitrogen supply (N2), the yellow colored solution was cooled to 10-15 °C. Next, 6.8 kg (54.7 mol) of oxalyl chloride were dosed in under N 2 over the course of 35 min from a supply container so that the temperature in the mixture did not rise above 20 °C.

After the addition was complete, the now dark green, thin suspension was stirred for 20-30 min at a mixture temperature of 18-25 °C.

In a 500 1 reactor, 18 kg of crushed ice was filled. The 25 °C hot suspension was added to the ice over 5 min so that the temperature of the mixture did not rise above 20 °C.

The reaction mixture was stirred for a further 10-20 min at a temperature in the mixture of 25-35 °C. The still green solution was diluted at 25-35 °C with 62.2 kg of diethylene glycol. Then, 14.9 kg (298 mol) of hydrazine hydrate were added from a supply vessel while cooling over the course of 10-15 minutes. The temperature in the mixture rose to a maximum of 40-45 °C. By gradually increasing the temperature over the course of 1.5 h, the now beige-colored suspension was heated to an internal temperature of 70-75 °C, whereby the THF distilled off. At an achieved internal temperature of 75 °C, 45.4 kg of THF distillate had been collected.

The reaction mixture was cooled to 50-55 °C and divided into 8 to 10 portions, with a total of 26.4 kg of potassium hydroxide (KOH) added in flakes over the course of 45 minutes, whereby the temperature of the mixture had already risen to 65- 70 °C and the initially thick suspension was colored yellow, became thinner and briefly had a slight reflux.

This suspension was now watered up with a temperature increase of 15 °C/h to 90 °C. From 85 °C, a slight foaming started and the suspension thickened. With a temperature rise of 2 °C/h, the temperature in the mixture was increased further to 102 °C, and at the same time as the stirrer speed was increased, nitrogen was blown through the reaction mixture by means of the dip tube. Due to strong foaming and, in addition, gas evolution, the volume of the reactor contents doubled. If necessary, the reaction temperature was lowered by means of cooling. At an internal temperature of 100-105 °C, the foam began to collapse and a reddish-brown, thin suspension was formed, which was further dewatered at a heating rate of 15 °C/h to an internal temperature of 140-145 °C. If foaming was too strong, the reaction temperature was briefly lowered by means of cooling. At the same time, several aqueous distillates totaling 44 kg were collected.

The batch was kept at 120-145 °C for 2-2.5 h. The reactor temperature was then lowered to 30-40 °C and 74.7 kg of water and 56.7 kg of diethyl ether were added. The reaction mixture was stirred for 10-15 min at an internal temperature of 30-33 °C, and the phases were then allowed to settle. The lower, strongly basic aqueous phase, weighing 154.9 kg, was colorless and only slightly turbid. It was removed as waste water. The yellow-colored, turbid intermediate phase with an oily consistency weighed 29.6 kg and contained the bulk of the product as potassium salt. The top, clear, yellow-coloured ethereal phase was vigorously stirred in an extraction apparatus with 10 kg of water for 10 min at a temperature of 30 °C in the mixture. 10 min after the end of stirring, the water phase was separated. The intermediate phase (29.6 kg) and the aqueous extract of the ether phase (10.9 kg) were mixed in an extraction apparatus with 126.2 kg of diethyl ether and 59.7 kg of water, and the mixture was cooled to an internal temperature of 0-5° C.

Via a supply vessel, a mixture of 6.0 kg of 32.5% hydrochloric acid and 6.0 kg of water was dosed in over the course of 15 minutes so that a maximum internal temperature of 10 °C was not exceeded and it was achieved a pH value of 1-2. If this pH value was not reached, a further 0.2 kg of 32.5% hydrochloric acid was added in a mixture with 0.2 kg of water. After this pH value was reached, the phases were stirred well for a further 5-10 min and then the stirring was stopped and the mixture was allowed to stand for 10-20 min for phase separation. Via the supply vessel, the ether phase was once again added to a mixture of 9.5 kg of hydrochloric acid and 19 kg of water and stirred well for 5-10 min at a temperature in the mixture of no more than 10 °C. The phases were separated and the HC1 treatment repeated up to three times if desired.

The ether phase was then mixed with 30 kg of demineralized water, stirred well for 10-20 min and heated to 15-20 °C. The phases were separated and the extraction repeated.

The ether phase was washed until it was free of traces of acid and added 6.5 kg of anhydrous magnesium sulfate and 0.4 kg of activated carbon ("Acticarbon 2S") which was slurried in 1 kg of diethyl ether, and stirred for 30-45 in at 18° C. The suspension was clearly filtered through a pressure filter coated with 0.5 kg of filtration aid (Cell flock) in a distillation apparatus. The filter and apparatus were rinsed with 8 kg of diethyl ether.

The ether phase was added to 95.6 kg of heptane, and at a temperature in the mixture of 15-20 °C

the ether was distilled off under vacuum. The formed crystal suspension after the distillation of ether was cooled to an internal temperature of 13-18 °C and stirred at this temperature for 0.5-1.5 h. The crystals were then centrifuged off. The moist product obtained was washed with 23.0 kg of heptane in 2 portions. The moist product was dried in a vacuum drying cabinet overnight at 50-60 °C, and if desired ground. 10.5 kg (77.2%) of ML-3000 was obtained with a melting point, determined according to the DSC method, of 157 °C. The IR spectrum was the same as the reference standard.

Claims (14)

1. Procedure for the preparation of the compound with fonnel I, characterized in that a) the compound of formula IV is converted into the compound of formula III by that al) the compound of formula IV is catalytically hydrogenated, or a2) the compound of formula IV is transferred to the ketal of formula IVa where the radicals R, which may be the same or different, are C-M-alkyl or together form C2.3-alMen> and the ketal is catalytically hydrogenated, b) the compound of formula III is reacted with co-bromo-4-chloroacetophenone to a compound of formula II and c) in the compound of formula II an acetic acid radical is introduced. 2. Process for preparing the compound of formula II characterized in that the compound of formula IV is converted to the compound of formula III by al) the compound of formula IV is catalytically hydrogenated, or a2) the compound of formula IV is converted to the ketal of formula IVa where the radicals R, which may be the same or different, are C1-alkyl or together form C2-3-alkylene, and the ketal is catalytically hydrogenated, and b) the compound of formula JII is reacted with co-bromo-4-chloroacetophenone to a compound of formula II. 3. Process for preparing the compound of formula III, characterized in that the compound of formula IV is converted to the compound of formula 111, by al) the compound of formula IV is catalytically hydrogenated, or a2) the compound of formula IV is converted into the ketal with formula IVa where the radicals R, which may be the same or different, are C1-4-alkyl or together form C2-3-alkylene, and the ketal is catalytically hydrogenated. 4. Method according to claims 1-3, where the catalytic hydrogenation is carried out with anhydrous Raney nickel as catalyst. 5. Process according to claims 1-4, where the hydrogenation is carried out in toluene or a mixture of toluene and a Ci_4 alcohol as solvent. 6. Method according to claims 1-5, where the compound of formula IV used has a purity of at least 95%. 7. Process according to claims 1-6, where the compound of formula IV is prepared by Michael addition of isobutyronitrile to a compound of formula V benzylation of the Michael addition product to 2-benzyl-4,4-dimethyl-2-(N-methylaniline)-glutarnitrile and hydrolysis of this nitrile. 8. Process according to claim 7, where the isobutyronitrile is deprotonated with lithium diisopropylamide in toluene. 9. Method according to claim 7 or 8, where the reaction temperature in the Michael addition lies in the range from -10 °C to -20 °C. 10. Process according to claims 7-9, where the hydrolysis of nitrile in acid is carried out in a two-phase system with phase transfer catalysis. 11. Process according to claims 7-10, where the compound of formula V is prepared by reaction of chloroacetaldehyde, N-methylaniline and an alkali metal cyanide, and subsequent basic elimination. 12. Method according to claim 11, where chloroacetaldehyde, and then alkali metal cyanide, are added to N-methylaniline. 13. Method according to claim 11 or 12, where chloroacetaldehyde, N-methylaniline and the alkali metal cyanide are used in a molar ratio of 1.1-1.3:1:1.1-1.3. 14. Method according to claims 11-13, where the basic elimination in the two-phase system is carried out during phase transfer catalysis.